Is there a way to trim values, with the same implementation as .replace()?

[FonixAU]

I am trying to do some dynamic flattening, and the only issue I cant get around, is that there are some characters I am replacing, that may appear at the ends and inside the string, so I am not able to stop my macro from creating trailing underscores

eg:{{ row.replace("．",'_') .replace(".",'_') .replace(" : ",'_') .replace(":",'') .replace(",",'_') .replace("?",'') .replace("-", '_') .replace('/', '_') .replace('(','_') .replace(')','') .replace(' ', '_') .replace('___','_') .replace('__','_').upper() }}

[jpgomezv]

Jinja is Python under the hood, so you can use Python string methods directly on the value. Add .rstrip('_') at the end of your chain — it removes any number of trailing underscores, which is exactly what your replace(...) calls leave behind:

{{ row.replace("．",'')
.replace(".",'')
.replace(" : ",'')
.replace(":",'')
.replace(",",'')
.replace("?",'')
.replace("-", '')
.replace('/', '')
.replace('(','')
.replace(')','')
.replace(' ', '')
.replace('','')
.replace('','')
.upper()
.rstrip('') }}

If leading underscores can also appear (e.g. the value starts with a replaced character), add .lstrip('_') as well. Note that rstrip/lstrip operate on the string as a whole, so they're safe to use after collapsing runs with .replace('___','_') / .replace('__','_').

If this is for generating column or alias names, an even cleaner option is the dbt_utils.slugify macro, which lowercases and collapses runs of non-alphanumeric characters into single underscores.

[jpgomezv]

Just a friendly follow-up — if my earlier answer helped resolve your issue, feel free to mark it as the accepted answer. Thanks!

[FonixAU]

Hi I havent worked in that team in over 2 years but I will let them know. Thank you.

…

On Tue, Aug 18, 2026 at 7:58 AM Juan Pablo Gómez Veira < ***@***.***> wrote: Just a friendly follow-up — if my earlier answer helped resolve your issue, feel free to mark it as the accepted answer. Thanks! — Reply to this email directly, view it on GitHub <#10821>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AM4DNLIEGHJUA4CJIEDJ4VD5KN5Z7AVCNFSNUABGKJSXA33TNF2G64TZHM2TGNJUHA4DMNZ3IRUXGY3VONZWS33OHM3TENZUGU2DJILWAI> . You are receiving this because you authored the thread.Message ID: ***@***.***>